SCOTUS-Stevens: Decision.fctr classification: caret Random Forest
bdanalytics
Date: (Thu) Apr 02, 2015
Introduction:
Data: Source: Training: https://courses.edx.org/c4x/MITx/15.071x_2/asset/stevens.csv
New:
Time period:
Synopsis:
Based on analysis utilizing <> techniques,
[](.png)
Potential next steps include:
Analysis:
rm(list=ls())
set.seed(12345)
options(stringsAsFactors=FALSE)
source("~/Dropbox/datascience/R/mydsutils.R")
source("~/Dropbox/datascience/R/myplot.R")
source("~/Dropbox/datascience/R/mypetrinet.R")
# Gather all package requirements here
#suppressPackageStartupMessages(require())
#require(sos); findFn("pinv", maxPages=2, sortby="MaxScore")
# Analysis control global variables
glb_is_separate_newent_dataset <- FALSE # or TRUE
glb_split_entity_newent_datasets <- TRUE # or FALSE
glb_split_newdata_method <- "sample" # "condition" or "sample"
glb_split_newdata_condition <- "<col_name> <condition_operator> <value>" # or NULL
glb_split_newdata_size_ratio <- 0.3 # > 0 & < 1
glb_split_sample.seed <- 3000 # or any integer
# for random forests (method="rf") & classification, the prediction variable has to be a factor
# for caret if the factor is based on numbers e.g (0/1 vs. "A"/"B"), caret prediction probabilities crashes
glb_predct_var <- "Decision.fctr" # or NULL
glb_predct_var_name <- paste0(glb_predct_var, ".predict")
glb_id_vars <- c("Docket") # or NULL
glb_exclude_vars_as_features <- union(glb_id_vars, ".rnorm") # or NULL
# List chrs (convert into factors if it's a valid feature); num/int transformed
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features,
c("Reverse", "Reverse.fctr",
"Circuit", "Issue", "Petitioner", "Respondent", "LowerCourt")
)
# List feats that shd be excluded due to known causation by prediction variable
glb_exclude_vars_as_features <- union(union(glb_exclude_vars_as_features,
paste(glb_predct_var_name, c("", ".proba"), sep="")),
c(NULL)) # or "<col_name>"
glb_impute_na_data <- FALSE # or TRUE
glb_mice_complete.seed <- 144 # or any integer
glb_is_regression <- FALSE; glb_is_classification <- TRUE
glb_sel_mdl <- glb_dmy_mdl <- NULL;
glb_models_method_vctr <- c("glm", "rpart", "rf")
glb_models_lst <- list(); glb_models_df <- data.frame()
glb_tune_models_df <- rbind(
data.frame(parameter="cp", min=0.01, max=0.5, by=0.01),
data.frame(parameter="mtry", min=2, max=4, by=1)) # or NULL
glb_clf_proba_threshold <- 0.5 # NULL
# Depict process
glb_analytics_pn <- petrinet(name="glb_analytics_pn",
trans_df=data.frame(id=1:6,
name=c("data.training.all","data.new",
"model.selected","model.final",
"data.training.all.prediction","data.new.prediction"),
x=c( -5,-5,-15,-25,-25,-35),
y=c( -5, 5, 0, 0, -5, 5)
),
places_df=data.frame(id=1:4,
name=c("bgn","fit.data.training.all","predict.data.new","end"),
x=c( -0, -20, -30, -40),
y=c( 0, 0, 0, 0),
M0=c( 3, 0, 0, 0)
),
arcs_df=data.frame(
begin=c("bgn","bgn","bgn",
"data.training.all","model.selected","fit.data.training.all",
"fit.data.training.all","model.final",
"data.new","predict.data.new",
"data.training.all.prediction","data.new.prediction"),
end =c("data.training.all","data.new","model.selected",
"fit.data.training.all","fit.data.training.all","model.final",
"data.training.all.prediction","predict.data.new",
"predict.data.new","data.new.prediction",
"end","end")
))
#print(ggplot.petrinet(glb_analytics_pn))
print(ggplot.petrinet(glb_analytics_pn) + coord_flip())## Loading required package: grid
glb_analytics_avl_objs <- NULL
glb_script_tm <- proc.time()
glb_script_df <- data.frame(chunk_label="import_data",
chunk_step_major=1, chunk_step_minor=0,
elapsed=(proc.time() - glb_script_tm)["elapsed"])
print(tail(glb_script_df, 2))## chunk_label chunk_step_major chunk_step_minor elapsed
## elapsed import_data 1 0 0.002Step 1: import data
glb_entity_df <- myimport_data(
url="https://courses.edx.org/c4x/MITx/15.071x_2/asset/stevens.csv",
comment="glb_entity_df", force_header=TRUE,
print_diagn=(glb_is_separate_newent_dataset |
!glb_split_entity_newent_datasets))## [1] "Reading file ./data/stevens.csv..."
## [1] "dimensions of data in ./data/stevens.csv: 566 rows x 9 cols"if (glb_is_separate_newent_dataset) {
glb_newent_df <- myimport_data(
url="<newdt_url>",
comment="glb_newent_df", force_header=TRUE, print_diagn=TRUE)
} else {
if (!glb_split_entity_newent_datasets) {
stop("Not implemented yet")
glb_newent_df <- glb_entity_df[sample(1:nrow(glb_entity_df),
max(2, nrow(glb_entity_df) / 1000)),]
} else if (glb_split_newdata_method == "condition") {
glb_newent_df <- do.call("subset",
list(glb_entity_df, parse(text=glb_split_newdata_condition)))
glb_entity_df <- do.call("subset",
list(glb_entity_df, parse(text=paste0("!(",
glb_split_newdata_condition,
")"))))
} else if (glb_split_newdata_method == "sample") {
require(caTools)
set.seed(glb_split_sample.seed)
split <- sample.split(glb_entity_df[, "Reverse"],
SplitRatio=(1-glb_split_newdata_size_ratio))
glb_newent_df <- glb_entity_df[!split, ]
glb_entity_df <- glb_entity_df[split ,]
} else stop("glb_split_newdata_method should be %in% c('condition', 'sample')")
comment(glb_newent_df) <- "glb_newent_df"
myprint_df(glb_newent_df)
str(glb_newent_df)
if (glb_split_entity_newent_datasets) {
myprint_df(glb_entity_df)
str(glb_entity_df)
}
} ## Loading required package: caTools## Docket Term Circuit Issue Petitioner Respondent
## 1 93-1408 1994 2nd EconomicActivity BUSINESS BUSINESS
## 3 93-1612 1994 5th EconomicActivity BUSINESS BUSINESS
## 4 94-623 1994 1st EconomicActivity BUSINESS BUSINESS
## 6 95-129 1995 9th EconomicActivity BUSINESS BUSINESS
## 8 96-1768 1997 9th EconomicActivity BUSINESS BUSINESS
## 21 97-1704 1998 5th JudicialPower INJURED.PERSON BUSINESS
## LowerCourt Unconst Reverse
## 1 liberal 0 1
## 3 liberal 0 1
## 4 conser 0 1
## 6 conser 1 0
## 8 conser 1 1
## 21 liberal 1 0
## Docket Term Circuit Issue Petitioner Respondent
## 187 99-536 1999 5th CivilRights EMPLOYEE EMPLOYER
## 188 00-6029 2001 8th Unions EMPLOYEE EMPLOYER
## 286 99-1030 2000 7th CriminalProcedure CITY OTHER
## 359 97-371 1997 9th FirstAmendment OTHER OTHER
## 452 95-6510 1995 4th DueProcess CRIMINAL.DEFENDENT STATE
## 484 99-9136 2000 9th CriminalProcedure CRIMINAL.DEFENDENT STATE
## LowerCourt Unconst Reverse
## 187 conser 0 1
## 188 liberal 0 0
## 286 liberal 0 0
## 359 liberal 0 1
## 452 liberal 1 0
## 484 conser 1 1
## Docket Term Circuit Issue Petitioner Respondent
## 545 00-507 2001 10th CivilRights AMERICAN.INDIAN US
## 546 95-1478 1996 9th CriminalProcedure GOVERNMENT.OFFICIAL US
## 551 00-157 2000 4th FederalTaxation GOVERNMENT.OFFICIAL US
## 552 00-1567 2001 1st EconomicActivity GOVERNMENT.OFFICIAL US
## 556 95-173 1995 9th EconomicActivity OTHER US
## 558 96-6839 1997 5th CriminalProcedure OTHER US
## LowerCourt Unconst Reverse
## 545 conser 0 0
## 546 liberal 1 0
## 551 liberal 0 0
## 552 liberal 0 0
## 556 conser 1 1
## 558 conser 1 1
## 'data.frame': 170 obs. of 9 variables:
## $ Docket : chr "93-1408" "93-1612" "94-623" "95-129" ...
## $ Term : int 1994 1994 1994 1995 1997 1998 1994 1995 1996 1996 ...
## $ Circuit : chr "2nd" "5th" "1st" "9th" ...
## $ Issue : chr "EconomicActivity" "EconomicActivity" "EconomicActivity" "EconomicActivity" ...
## $ Petitioner: chr "BUSINESS" "BUSINESS" "BUSINESS" "BUSINESS" ...
## $ Respondent: chr "BUSINESS" "BUSINESS" "BUSINESS" "BUSINESS" ...
## $ LowerCourt: chr "liberal" "liberal" "conser" "conser" ...
## $ Unconst : int 0 0 0 1 1 1 0 0 1 0 ...
## $ Reverse : int 1 1 1 0 1 0 0 0 0 1 ...
## - attr(*, "comment")= chr "glb_newent_df"
## Docket Term Circuit Issue Petitioner Respondent LowerCourt
## 2 93-1577 1994 9th EconomicActivity BUSINESS BUSINESS liberal
## 5 94-1175 1995 7th JudicialPower BUSINESS BUSINESS conser
## 7 95-728 1996 FED EconomicActivity BUSINESS BUSINESS conser
## 9 96-843 1997 DC EconomicActivity BUSINESS BUSINESS conser
## 10 98-1480 1999 11th CriminalProcedure BUSINESS BUSINESS conser
## 11 99-150 1999 2nd EconomicActivity BUSINESS BUSINESS conser
## Unconst Reverse
## 2 0 1
## 5 0 1
## 7 0 1
## 9 0 1
## 10 0 1
## 11 0 1
## Docket Term Circuit Issue Petitioner
## 20 95-1872 1996 8th CivilRights AMERICAN.INDIAN
## 49 96-1971 1997 5th CriminalProcedure OTHER
## 52 97-704 1997 DC JudicialPower OTHER
## 73 94-1893 1995 4th FirstAmendment US
## 232 94-500 1994 5th FederalTaxation OTHER
## 401 95-489 1995 10th FirstAmendment POLITICIAN
## Respondent LowerCourt Unconst Reverse
## 20 BUSINESS conser 0 0
## 49 BUSINESS liberal 1 1
## 52 BUSINESS conser 0 0
## 73 BUSINESS liberal 0 1
## 232 GOVERNMENT.OFFICIAL conser 0 1
## 401 OTHER conser 1 0
## Docket Term Circuit Issue Petitioner Respondent
## 561 98-1828 1999 2nd CriminalProcedure OTHER US
## 562 99-5153 1999 6th CriminalProcedure OTHER US
## 563 99-804 2000 5th CriminalProcedure OTHER US
## 564 99-8508 2000 9th CriminalProcedure OTHER US
## 565 97-29 1997 DC CivilRights STATE US
## 566 00-189 2000 9th CivilRights STATE US
## LowerCourt Unconst Reverse
## 561 conser 1 1
## 562 conser 0 0
## 563 liberal 0 1
## 564 conser 1 0
## 565 conser 0 0
## 566 liberal 0 0
## 'data.frame': 396 obs. of 9 variables:
## $ Docket : chr "93-1577" "94-1175" "95-728" "96-843" ...
## $ Term : int 1994 1995 1996 1997 1999 1999 2000 2000 1994 1999 ...
## $ Circuit : chr "9th" "7th" "FED" "DC" ...
## $ Issue : chr "EconomicActivity" "JudicialPower" "EconomicActivity" "EconomicActivity" ...
## $ Petitioner: chr "BUSINESS" "BUSINESS" "BUSINESS" "BUSINESS" ...
## $ Respondent: chr "BUSINESS" "BUSINESS" "BUSINESS" "BUSINESS" ...
## $ LowerCourt: chr "liberal" "conser" "conser" "conser" ...
## $ Unconst : int 0 0 0 0 0 0 1 1 0 0 ...
## $ Reverse : int 1 1 1 1 1 1 1 1 0 0 ...
## - attr(*, "comment")= chr "glb_entity_df"glb_script_df <- rbind(glb_script_df,
data.frame(chunk_label="cleanse_data",
chunk_step_major=max(glb_script_df$chunk_step_major)+1,
chunk_step_minor=0,
elapsed=(proc.time() - glb_script_tm)["elapsed"]))
print(tail(glb_script_df, 2))## chunk_label chunk_step_major chunk_step_minor elapsed
## elapsed import_data 1 0 0.002
## elapsed1 cleanse_data 2 0 0.564Step 2: cleanse data
glb_script_df <- rbind(glb_script_df,
data.frame(chunk_label="inspect_explore_data",
chunk_step_major=max(glb_script_df$chunk_step_major),
chunk_step_minor=1,
elapsed=(proc.time() - glb_script_tm)["elapsed"]))
print(tail(glb_script_df, 2))## chunk_label chunk_step_major chunk_step_minor elapsed
## elapsed1 cleanse_data 2 0 0.564
## elapsed2 inspect_explore_data 2 1 0.595Step 2.1: inspect/explore data
#print(str(glb_entity_df))
#View(glb_entity_df)
# List info gathered for various columns
# <col_name>: <description>; <notes>
# Create new features that help diagnostics
# Convert factors to dummy variables
# Build splines require(splines); bsBasis <- bs(training$age, df=3)
add_new_diag_feats <- function(obs_df, obs_twin_df) {
require(plyr)
obs_df <- mutate(obs_df,
# <col_name>.NA=is.na(<col_name>),
Reverse.fctr=relevel(factor(Reverse,
as.factor(union(obs_df$Reverse, obs_twin_df$Reverse))),
ref="0"),
Decision.fctr=relevel(factor(ifelse(Reverse == 1, "R", "A")),
as.factor(c("R", "A")),
ref="A"),
Circuit.fctr=factor(Circuit,
as.factor(union(obs_df$Circuit, obs_twin_df$Circuit))),
Issue.fctr=factor(Issue,
as.factor(union(obs_df$Issue, obs_twin_df$Issue))),
Petitioner.fctr=factor(Petitioner,
as.factor(union(obs_df$Petitioner, obs_twin_df$Petitioner))),
Respondent.fctr=factor(Respondent,
as.factor(union(obs_df$Respondent, obs_twin_df$Respondent))),
LowerCourt.fctr=factor(LowerCourt,
as.factor(union(obs_df$LowerCourt, obs_twin_df$LowerCourt))),
# <col_name>.fctr=factor(<col_name>,
# as.factor(union(obs_df$<col_name>, obs_twin_df$<col_name>))),
# <col_name>.fctr=relevel(factor(<col_name>,
# as.factor(union(obs_df$<col_name>, obs_twin_df$<col_name>))),
# "<ref_val>"),
# <col2_name>.fctr=relevel(factor(ifelse(<col1_name> == <val>, "<oth_val>", "<ref_val>")),
# as.factor(c("R", "<ref_val>")),
# ref="<ref_val>"),
# This doesn't work - use sapply instead
# <col_name>.fctr_num=grep(<col_name>, levels(<col_name>.fctr)),
#
# Date.my=as.Date(strptime(Date, "%m/%d/%y %H:%M")),
# Year=year(Date.my),
# Month=months(Date.my),
# Weekday=weekdays(Date.my)
# <col_name>.log=log(<col.name>),
# <col_name>=<table>[as.character(<col2_name>)],
# <col_name>=as.numeric(<col2_name>),
.rnorm=rnorm(n=nrow(obs_df))
)
# If levels of a factor are different across obs_df & glb_newent_df; predict.glm fails
# Transformations not handled by mutate
# obs_df$<col_name>.fctr.num <- sapply(1:nrow(obs_df),
# function(row_ix) grep(obs_df[row_ix, "<col_name>"],
# levels(obs_df[row_ix, "<col_name>.fctr"])))
print(summary(obs_df))
print(sapply(names(obs_df), function(col) sum(is.na(obs_df[, col]))))
return(obs_df)
}
glb_entity_df <- add_new_diag_feats(glb_entity_df, glb_newent_df)## Loading required package: plyr## Docket Term Circuit Issue
## Length:396 Min. :1994 Length:396 Length:396
## Class :character 1st Qu.:1995 Class :character Class :character
## Mode :character Median :1997 Mode :character Mode :character
## Mean :1997
## 3rd Qu.:1999
## Max. :2001
##
## Petitioner Respondent LowerCourt Unconst
## Length:396 Length:396 Length:396 Min. :0.0000
## Class :character Class :character Class :character 1st Qu.:0.0000
## Mode :character Mode :character Mode :character Median :0.0000
## Mean :0.2247
## 3rd Qu.:0.0000
## Max. :1.0000
##
## Reverse Reverse.fctr Decision.fctr Circuit.fctr
## Min. :0.0000 0:180 A:180 9th : 79
## 1st Qu.:0.0000 1:216 R:216 11th : 41
## Median :1.0000 5th : 40
## Mean :0.5455 4th : 33
## 3rd Qu.:1.0000 8th : 32
## Max. :1.0000 7th : 31
## (Other):140
## Issue.fctr Petitioner.fctr
## CriminalProcedure:92 OTHER :123
## JudicialPower :78 CRIMINAL.DEFENDENT : 58
## EconomicActivity :58 BUSINESS : 53
## CivilRights :55 STATE : 38
## DueProcess :36 US : 34
## FirstAmendment :25 GOVERNMENT.OFFICIAL: 25
## (Other) :52 (Other) : 65
## Respondent.fctr LowerCourt.fctr .rnorm
## OTHER :120 liberal:185 Min. :-2.74448
## BUSINESS : 58 conser :211 1st Qu.:-0.58477
## US : 52 Median :-0.04916
## CRIMINAL.DEFENDENT: 45 Mean : 0.07455
## STATE : 35 3rd Qu.: 0.80318
## EMPLOYEE : 18 Max. : 3.06873
## (Other) : 68
## Docket Term Circuit Issue
## 0 0 0 0
## Petitioner Respondent LowerCourt Unconst
## 0 0 0 0
## Reverse Reverse.fctr Decision.fctr Circuit.fctr
## 0 0 0 0
## Issue.fctr Petitioner.fctr Respondent.fctr LowerCourt.fctr
## 0 0 0 0
## .rnorm
## 0glb_newent_df <- add_new_diag_feats(glb_newent_df, glb_entity_df)## Docket Term Circuit Issue
## Length:170 Min. :1994 Length:170 Length:170
## Class :character 1st Qu.:1995 Class :character Class :character
## Mode :character Median :1997 Mode :character Mode :character
## Mean :1997
## 3rd Qu.:1999
## Max. :2001
##
## Petitioner Respondent LowerCourt Unconst
## Length:170 Length:170 Length:170 Min. :0.0
## Class :character Class :character Class :character 1st Qu.:0.0
## Mode :character Mode :character Mode :character Median :0.0
## Mean :0.3
## 3rd Qu.:1.0
## Max. :1.0
##
## Reverse Reverse.fctr Decision.fctr Circuit.fctr
## Min. :0.0000 0:77 A:77 9th :43
## 1st Qu.:0.0000 1:93 R:93 7th :16
## Median :1.0000 6th :14
## Mean :0.5471 5th :13
## 3rd Qu.:1.0000 4th :13
## Max. :1.0000 3rd :12
## (Other):59
## Issue.fctr Petitioner.fctr
## EconomicActivity :40 OTHER :52
## CriminalProcedure :40 CRIMINAL.DEFENDENT :31
## JudicialPower :24 BUSINESS :26
## CivilRights :19 US :14
## FirstAmendment :14 GOVERNMENT.OFFICIAL:13
## FederalismAndInterstateRelations:10 STATE :10
## (Other) :23 (Other) :24
## Respondent.fctr LowerCourt.fctr .rnorm
## OTHER :57 liberal:88 Min. :-3.0077
## BUSINESS :22 conser :82 1st Qu.:-0.6076
## STATE :21 Median : 0.1553
## US :17 Mean : 0.1304
## CRIMINAL.DEFENDENT:13 3rd Qu.: 0.8110
## EMPLOYEE :10 Max. : 3.5577
## (Other) :30
## Docket Term Circuit Issue
## 0 0 0 0
## Petitioner Respondent LowerCourt Unconst
## 0 0 0 0
## Reverse Reverse.fctr Decision.fctr Circuit.fctr
## 0 0 0 0
## Issue.fctr Petitioner.fctr Respondent.fctr LowerCourt.fctr
## 0 0 0 0
## .rnorm
## 0# Histogram of predictor in glb_entity_df & glb_newent_df
print(myplot_histogram(glb_entity_df, glb_predct_var))
#pairs(subset(glb_entity_df, select=-c(col_symbol)))
# Check for glb_newent_df & glb_entity_df features range mismatches
# Other diagnostics:
# print(subset(glb_entity_df, <col1_name> == max(glb_entity_df$<col1_name>, na.rm=TRUE) &
# <col2_name> <= mean(glb_entity_df$<col1_name>, na.rm=TRUE)))
# print(glb_entity_df[which.max(glb_entity_df$<col_name>),])
# print(<col_name>_freq_glb_entity_df <- mycreate_tbl_df(glb_entity_df, "<col_name>"))
# print(which.min(table(glb_entity_df$<col_name>)))
# print(which.max(table(glb_entity_df$<col_name>)))
# print(which.max(table(glb_entity_df$<col1_name>, glb_entity_df$<col2_name>)[, 2]))
# print(table(glb_entity_df$<col1_name>, glb_entity_df$<col2_name>))
# print(table(is.na(glb_entity_df$<col1_name>), glb_entity_df$<col2_name>))
# print(table(sign(glb_entity_df$<col1_name>), glb_entity_df$<col2_name>))
# print(mycreate_xtab(glb_entity_df, <col1_name>))
# print(mycreate_xtab(glb_entity_df, c(<col1_name>, <col2_name>)))
# print(<col1_name>_<col2_name>_xtab_glb_entity_df <-
# mycreate_xtab(glb_entity_df, c("<col1_name>", "<col2_name>")))
# <col1_name>_<col2_name>_xtab_glb_entity_df[is.na(<col1_name>_<col2_name>_xtab_glb_entity_df)] <- 0
# print(<col1_name>_<col2_name>_xtab_glb_entity_df <-
# mutate(<col1_name>_<col2_name>_xtab_glb_entity_df,
# <col3_name>=(<col1_name> * 1.0) / (<col1_name> + <col2_name>)))
# print(<col2_name>_min_entity_arr <-
# sort(tapply(glb_entity_df$<col1_name>, glb_entity_df$<col2_name>, min, na.rm=TRUE)))
# print(<col1_name>_na_by_<col2_name>_arr <-
# sort(tapply(glb_entity_df$<col1_name>.NA, glb_entity_df$<col2_name>, mean, na.rm=TRUE)))
# Other plots:
# print(myplot_box(df=glb_entity_df, ycol_names="<col1_name>"))
# print(myplot_box(df=glb_entity_df, ycol_names="<col1_name>", xcol_name="<col2_name>"))
# print(myplot_line(subset(glb_entity_df, Symbol %in% c("KO", "PG")),
# "Date.my", "StockPrice", facet_row_colnames="Symbol") +
# geom_vline(xintercept=as.numeric(as.Date("2003-03-01"))) +
# geom_vline(xintercept=as.numeric(as.Date("1983-01-01")))
# )
# print(myplot_scatter(glb_entity_df, "<col1_name>", "<col2_name>", smooth=TRUE))
# print(myplot_scatter(glb_entity_df, "<col1_name>", "<col2_name>", colorcol_name="<Pred.fctr>"))
glb_script_df <- rbind(glb_script_df,
data.frame(chunk_label="manage_missing_data",
chunk_step_major=max(glb_script_df$chunk_step_major),
chunk_step_minor=glb_script_df[nrow(glb_script_df), "chunk_step_minor"]+1,
elapsed=(proc.time() - glb_script_tm)["elapsed"]))
print(tail(glb_script_df, 2))## chunk_label chunk_step_major chunk_step_minor elapsed
## elapsed2 inspect_explore_data 2 1 0.595
## elapsed3 manage_missing_data 2 2 1.428Step 2.2: manage missing data
# print(sapply(names(glb_entity_df), function(col) sum(is.na(glb_entity_df[, col]))))
# print(sapply(names(glb_newent_df), function(col) sum(is.na(glb_newent_df[, col]))))
# glb_entity_df <- na.omit(glb_entity_df)
# glb_newent_df <- na.omit(glb_newent_df)
# df[is.na(df)] <- 0
# Not refactored into mydsutils.R since glb_*_df might be reassigned
glb_impute_missing_data <- function(entity_df, newent_df) {
if (!glb_is_separate_newent_dataset) {
# Combine entity & newent
union_df <- rbind(mutate(entity_df, .src = "entity"),
mutate(newent_df, .src = "newent"))
union_imputed_df <- union_df[, setdiff(setdiff(names(entity_df),
glb_predct_var),
glb_exclude_vars_as_features)]
print(summary(union_imputed_df))
require(mice)
set.seed(glb_mice_complete.seed)
union_imputed_df <- complete(mice(union_imputed_df))
print(summary(union_imputed_df))
union_df[, names(union_imputed_df)] <- union_imputed_df[, names(union_imputed_df)]
print(summary(union_df))
# union_df$.rownames <- rownames(union_df)
# union_df <- orderBy(~.rownames, union_df)
#
# imp_entity_df <- myimport_data(
# url="<imputed_trnng_url>",
# comment="imp_entity_df", force_header=TRUE, print_diagn=TRUE)
# print(all.equal(subset(union_df, select=-c(.src, .rownames, .rnorm)),
# imp_entity_df))
# Partition again
glb_entity_df <<- subset(union_df, .src == "entity", select=-c(.src, .rownames))
comment(glb_entity_df) <- "entity_df"
glb_newent_df <<- subset(union_df, .src == "newent", select=-c(.src, .rownames))
comment(glb_newent_df) <- "newent_df"
# Generate summaries
print(summary(entity_df))
print(sapply(names(entity_df), function(col) sum(is.na(entity_df[, col]))))
print(summary(newent_df))
print(sapply(names(newent_df), function(col) sum(is.na(newent_df[, col]))))
} else stop("Not implemented yet")
}
if (glb_impute_na_data) {
if ((sum(sapply(names(glb_entity_df),
function(col) sum(is.na(glb_entity_df[, col])))) > 0) |
(sum(sapply(names(glb_newent_df),
function(col) sum(is.na(glb_newent_df[, col])))) > 0))
glb_impute_missing_data(glb_entity_df, glb_newent_df)
}
glb_script_df <- rbind(glb_script_df,
data.frame(chunk_label="encode_retype_data",
chunk_step_major=max(glb_script_df$chunk_step_major),
chunk_step_minor=glb_script_df[nrow(glb_script_df), "chunk_step_minor"]+1,
elapsed=(proc.time() - glb_script_tm)["elapsed"]))
print(tail(glb_script_df, 2))## chunk_label chunk_step_major chunk_step_minor elapsed
## elapsed3 manage_missing_data 2 2 1.428
## elapsed4 encode_retype_data 2 3 1.668Step 2.3: encode/retype data
# map_<col_name>_df <- myimport_data(
# url="<map_url>",
# comment="map_<col_name>_df", print_diagn=TRUE)
# map_<col_name>_df <- read.csv(paste0(getwd(), "/data/<file_name>.csv"), strip.white=TRUE)
# glb_entity_df <- mymap_codes(glb_entity_df, "<from_col_name>", "<to_col_name>",
# map_<to_col_name>_df, map_join_col_name="<map_join_col_name>",
# map_tgt_col_name="<to_col_name>")
# glb_newent_df <- mymap_codes(glb_newent_df, "<from_col_name>", "<to_col_name>",
# map_<to_col_name>_df, map_join_col_name="<map_join_col_name>",
# map_tgt_col_name="<to_col_name>")
# glb_entity_df$<col_name>.fctr <- factor(glb_entity_df$<col_name>,
# as.factor(union(glb_entity_df$<col_name>, glb_newent_df$<col_name>)))
# glb_newent_df$<col_name>.fctr <- factor(glb_newent_df$<col_name>,
# as.factor(union(glb_entity_df$<col_name>, glb_newent_df$<col_name>)))
glb_script_df <- rbind(glb_script_df,
data.frame(chunk_label="extract_features",
chunk_step_major=max(glb_script_df$chunk_step_major)+1,
chunk_step_minor=0,
elapsed=(proc.time() - glb_script_tm)["elapsed"]))
print(tail(glb_script_df, 2))## chunk_label chunk_step_major chunk_step_minor elapsed
## elapsed4 encode_retype_data 2 3 1.668
## elapsed5 extract_features 3 0 1.715Step 3: extract features
# Create new features that help prediction
# <col_name>.lag.2 <- lag(zoo(glb_entity_df$<col_name>), -2, na.pad=TRUE)
# glb_entity_df[, "<col_name>.lag.2"] <- coredata(<col_name>.lag.2)
# <col_name>.lag.2 <- lag(zoo(glb_newent_df$<col_name>), -2, na.pad=TRUE)
# glb_newent_df[, "<col_name>.lag.2"] <- coredata(<col_name>.lag.2)
#
# glb_newent_df[1, "<col_name>.lag.2"] <- glb_entity_df[nrow(glb_entity_df) - 1,
# "<col_name>"]
# glb_newent_df[2, "<col_name>.lag.2"] <- glb_entity_df[nrow(glb_entity_df),
# "<col_name>"]
# glb_entity_df <- mutate(glb_entity_df,
# <new_col_name>=
# )
# glb_newent_df <- mutate(glb_newent_df,
# <new_col_name>=
# )
# print(summary(glb_entity_df))
# print(summary(glb_newent_df))
# print(sapply(names(glb_entity_df), function(col) sum(is.na(glb_entity_df[, col]))))
# print(sapply(names(glb_newent_df), function(col) sum(is.na(glb_newent_df[, col]))))
# print(myplot_scatter(glb_entity_df, "<col1_name>", "<col2_name>", smooth=TRUE))
replay.petrisim(pn=glb_analytics_pn,
replay.trans=(glb_analytics_avl_objs <- c(glb_analytics_avl_objs,
"data.training.all","data.new")), flip_coord=TRUE)## time trans "bgn " "fit.data.training.all " "predict.data.new " "end "
## 0.0000 multiple enabled transitions: data.training.all data.new model.selected firing: data.training.all
## 1.0000 1 2 1 0 0
## 1.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction firing: data.new
## 2.0000 2 1 1 1 0
glb_script_df <- rbind(glb_script_df,
data.frame(chunk_label="select_features",
chunk_step_major=max(glb_script_df$chunk_step_major)+1,
chunk_step_minor=0,
elapsed=(proc.time() - glb_script_tm)["elapsed"]))
print(tail(glb_script_df, 2))## chunk_label chunk_step_major chunk_step_minor elapsed
## elapsed5 extract_features 3 0 1.715
## elapsed6 select_features 4 0 2.526Step 4: select features
print(glb_feats_df <-
myselect_features(lcl_entity_df=glb_entity_df,
lcl_exclude_vars_as_features=glb_exclude_vars_as_features,
lcl_predct_var=glb_predct_var))## id cor.y cor.y.abs
## Unconst Unconst 0.06627144 0.06627144
## Term Term 0.03032304 0.03032304glb_script_df <- rbind(glb_script_df,
data.frame(chunk_label="remove_correlated_features",
chunk_step_major=max(glb_script_df$chunk_step_major),
chunk_step_minor=glb_script_df[nrow(glb_script_df), "chunk_step_minor"]+1,
elapsed=(proc.time() - glb_script_tm)["elapsed"]))
print(tail(glb_script_df, 2))## chunk_label chunk_step_major chunk_step_minor
## elapsed6 select_features 4 0
## elapsed7 remove_correlated_features 4 1
## elapsed
## elapsed6 2.526
## elapsed7 2.710Step 5: run models
max_cor_y_x_var <- subset(glb_feats_df, cor.low == 1)[1, "id"]
# Regression:
if (glb_is_regression) {
# Linear:
myrun_mdl_fn <- myrun_mdl_lm
}
# Classification:
if (glb_is_classification) myrun_mdl_fn <- myrun_mdl_classification
# Add dummy model - random variable
# Potential Enhancements:
# For classifiers, it shd generate proba/outcomes that mimics the freq
# distribution of glb_predct_var values; Right now it always generates
# 0 (most frequent ?)
ret_lst <- myrun_mdl_fn(indep_vars_vctr=".rnorm",
lcl_predct_var=glb_predct_var,
lcl_predct_var_name=glb_predct_var_name,
fit_df=glb_entity_df, OOB_df=glb_newent_df,
method="glm")## Loading required package: ROCR
## Loading required package: gplots
##
## Attaching package: 'gplots'
##
## The following object is masked from 'package:stats':
##
## lowess
##
## Loading required package: caret
## Loading required package: lattice
##
## Attaching package: 'caret'
##
## The following object is masked from 'package:survival':
##
## cluster
##
## Call: NULL
##
## Coefficients:
## (Intercept) .rnorm
## 0.18489 -0.03377
##
## Degrees of Freedom: 395 Total (i.e. Null); 394 Residual
## Null Deviance: 545.7
## Residual Deviance: 545.6 AIC: 549.6glb_dmy_mdl <- ret_lst[["model"]]
# Highest cor.y
ret_lst <- myrun_mdl_fn(indep_vars_vctr=max_cor_y_x_var,
lcl_predct_var=glb_predct_var,
lcl_predct_var_name=glb_predct_var_name,
fit_df=glb_entity_df, OOB_df=glb_newent_df,
method="glm")
##
## Call: NULL
##
## Coefficients:
## (Intercept) Unconst
## 0.1109 0.3228
##
## Degrees of Freedom: 395 Total (i.e. Null); 394 Residual
## Null Deviance: 545.7
## Residual Deviance: 543.9 AIC: 547.9# Enhance Highest cor.y model with additions of interaction terms that were
# dropped due to high correlations
if (nrow(subset(glb_feats_df, is.na(cor.low))) > 0)
ret_lst <- myrun_mdl_fn(indep_vars_vctr=c(max_cor_y_x_var,
paste(max_cor_y_x_var,
subset(glb_feats_df, is.na(cor.low))[, "id"], sep=":")),
glb_predct_var, glb_predct_var_name,
fit_df=glb_entity_df, OOB_df=glb_newent_df,
method="glm")
# Low correlated X
ret_lst <- myrun_mdl_fn(indep_vars_vctr=subset(glb_feats_df,
cor.low == 1)[, "id"],
glb_predct_var, glb_predct_var_name,
fit_df=glb_entity_df, OOB_df=glb_newent_df,
method="glm")
##
## Call: NULL
##
## Coefficients:
## (Intercept) Unconst Term
## -57.42477 0.32271 0.02881
##
## Degrees of Freedom: 395 Total (i.e. Null); 393 Residual
## Null Deviance: 545.7
## Residual Deviance: 543.6 AIC: 549.6# All X that is not user excluded
ret_lst <- myrun_mdl_fn(indep_vars_vctr=setdiff(names(glb_entity_df),
union(glb_predct_var, glb_exclude_vars_as_features)),
glb_predct_var, glb_predct_var_name,
fit_df=glb_entity_df, OOB_df=glb_newent_df,
method="glm")## Warning in predict.lm(object, newdata, se.fit, scale = 1, type =
## ifelse(type == : prediction from a rank-deficient fit may be misleading
##
## Call: NULL
##
## Coefficients:
## (Intercept)
## -47.91172
## Term
## 0.02433
## Unconst
## -0.17555
## Circuit.fctr7th
## -0.65021
## Circuit.fctrFED
## -0.72046
## Circuit.fctrDC
## -0.76550
## Circuit.fctr11th
## -0.23589
## Circuit.fctr2nd
## 0.93630
## Circuit.fctr6th
## 0.13244
## Circuit.fctr3rd
## -0.90093
## Circuit.fctr10th
## -1.35086
## Circuit.fctr8th
## -0.56162
## Circuit.fctr5th
## 0.68012
## Circuit.fctr4th
## 0.59484
## Circuit.fctr1st
## -0.87873
## Issue.fctrJudicialPower
## -0.32049
## Issue.fctrCriminalProcedure
## -0.30717
## Issue.fctrDueProcess
## 0.25310
## Issue.fctrFederalismAndInterstateRelations
## 0.08384
## Issue.fctrCivilRights
## -0.07564
## Issue.fctrFirstAmendment
## -0.73942
## Issue.fctrFederalTaxation
## -1.95641
## Issue.fctrUnions
## -1.01470
## Issue.fctrPrivacy
## 2.12821
## Issue.fctrAttorneys
## -0.21996
## Petitioner.fctrCITY
## -1.54739
## Petitioner.fctrEMPLOYEE
## 0.72629
## Petitioner.fctrAMERICAN.INDIAN
## -1.30924
## Petitioner.fctrINJURED.PERSON
## 0.08859
## Petitioner.fctrGOVERNMENT.OFFICIAL
## -1.00691
## Petitioner.fctrOTHER
## 0.05366
## Petitioner.fctrSTATE
## -0.54022
## Petitioner.fctrUS
## 0.11006
## Petitioner.fctrCRIMINAL.DEFENDENT
## 0.67585
## Petitioner.fctrEMPLOYER
## -0.64909
## Petitioner.fctrPOLITICIAN
## -0.59568
## Respondent.fctrCITY
## -0.86937
## Respondent.fctrCRIMINAL.DEFENDENT
## -1.27037
## Respondent.fctrEMPLOYEE
## -0.58328
## Respondent.fctrEMPLOYER
## 0.83259
## Respondent.fctrAMERICAN.INDIAN
## 1.74125
## Respondent.fctrINJURED.PERSON
## -1.71043
## Respondent.fctrGOVERNMENT.OFFICIAL
## -0.62522
## Respondent.fctrOTHER
## -0.49674
## Respondent.fctrPOLITICIAN
## 0.08058
## Respondent.fctrSTATE
## 0.38060
## Respondent.fctrUS
## -1.29052
## LowerCourt.fctrconser
## 0.96186
##
## Degrees of Freedom: 395 Total (i.e. Null); 348 Residual
## Null Deviance: 545.7
## Residual Deviance: 428.5 AIC: 524.5# User specified
for (method in glb_models_method_vctr) {
#print(sprintf("iterating over method:%s", method))
# easier to exclude features
indep_vars_vctr <- setdiff(names(glb_entity_df),
union(union(glb_predct_var, glb_exclude_vars_as_features),
c("Term")))
# easier to include features
# indep_vars_vctr <- c("<feat1_name>", "<feat2_name>")
# User specified bivariate models
# indep_vars_vctr_lst <- list()
# for (feat in setdiff(names(glb_entity_df),
# union(glb_predct_var, glb_exclude_vars_as_features)))
# indep_vars_vctr_lst[["feat"]] <- feat
# User specified combinatorial models
# indep_vars_vctr_lst <- list()
# combn_mtrx <- combn(c("<feat1_name>", "<feat2_name>", "<featn_name>"),
# <num_feats_to_choose>)
# for (combn_ix in 1:ncol(combn_mtrx))
# #print(combn_mtrx[, combn_ix])
# indep_vars_vctr_lst[[combn_ix]] <- combn_mtrx[, combn_ix]
set.seed(200)
ret_lst <- myrun_mdl_fn(indep_vars_vctr=indep_vars_vctr,
glb_predct_var, glb_predct_var_name,
fit_df=glb_entity_df, OOB_df=glb_newent_df,
method=method)
# glb_<id>_mdl <- ret_lst[["model"]]
}
##
## Call: NULL
##
## Coefficients:
## (Intercept)
## 0.66858
## Unconst
## -0.18029
## Circuit.fctr7th
## -0.64115
## Circuit.fctrFED
## -0.70973
## Circuit.fctrDC
## -0.76663
## Circuit.fctr11th
## -0.22491
## Circuit.fctr2nd
## 0.97359
## Circuit.fctr6th
## 0.14258
## Circuit.fctr3rd
## -0.88092
## Circuit.fctr10th
## -1.34112
## Circuit.fctr8th
## -0.55667
## Circuit.fctr5th
## 0.69556
## Circuit.fctr4th
## 0.61347
## Circuit.fctr1st
## -0.88244
## Issue.fctrJudicialPower
## -0.32201
## Issue.fctrCriminalProcedure
## -0.29264
## Issue.fctrDueProcess
## 0.27264
## Issue.fctrFederalismAndInterstateRelations
## 0.06834
## Issue.fctrCivilRights
## -0.05890
## Issue.fctrFirstAmendment
## -0.71501
## Issue.fctrFederalTaxation
## -1.98079
## Issue.fctrUnions
## -0.99795
## Issue.fctrPrivacy
## 2.18674
## Issue.fctrAttorneys
## -0.21192
## Petitioner.fctrCITY
## -1.55222
## Petitioner.fctrEMPLOYEE
## 0.73154
## Petitioner.fctrAMERICAN.INDIAN
## -1.36136
## Petitioner.fctrINJURED.PERSON
## 0.10632
## Petitioner.fctrGOVERNMENT.OFFICIAL
## -1.00998
## Petitioner.fctrOTHER
## 0.04596
## Petitioner.fctrSTATE
## -0.54067
## Petitioner.fctrUS
## 0.09438
## Petitioner.fctrCRIMINAL.DEFENDENT
## 0.65127
## Petitioner.fctrEMPLOYER
## -0.67638
## Petitioner.fctrPOLITICIAN
## -0.58921
## Respondent.fctrCITY
## -0.89541
## Respondent.fctrCRIMINAL.DEFENDENT
## -1.27049
## Respondent.fctrEMPLOYEE
## -0.58380
## Respondent.fctrEMPLOYER
## 0.81539
## Respondent.fctrAMERICAN.INDIAN
## 1.73754
## Respondent.fctrINJURED.PERSON
## -1.72998
## Respondent.fctrGOVERNMENT.OFFICIAL
## -0.64414
## Respondent.fctrOTHER
## -0.51043
## Respondent.fctrPOLITICIAN
## 0.06309
## Respondent.fctrSTATE
## 0.36823
## Respondent.fctrUS
## -1.29002
## LowerCourt.fctrconser
## 0.95835
##
## Degrees of Freedom: 395 Total (i.e. Null); 349 Residual
## Null Deviance: 545.7
## Residual Deviance: 428.7 AIC: 522.7## Loading required package: rpart
## Loading required package: rpart.plot
## n= 396
##
## node), split, n, loss, yval, (yprob)
## * denotes terminal node
##
## 1) root 396 180 R (0.4545455 0.5454545)
## 2) LowerCourt.fctrconser< 0.5 185 73 A (0.6054054 0.3945946)
## 4) Respondent.fctrAMERICAN.INDIAN< 0.5 175 65 A (0.6285714 0.3714286) *
## 5) Respondent.fctrAMERICAN.INDIAN>=0.5 10 2 R (0.2000000 0.8000000) *
## 3) LowerCourt.fctrconser>=0.5 211 68 R (0.3222749 0.6777251) *## Loading required package: randomForest
## randomForest 4.6-10
## Type rfNews() to see new features/changes/bug fixes.
##
## Call:
## randomForest(x = x, y = y, mtry = param$mtry)
## Type of random forest: classification
## Number of trees: 500
## No. of variables tried at each split: 2
##
## OOB estimate of error rate: 36.11%
## Confusion matrix:
## A R class.error
## A 81 99 0.5500000
## R 44 172 0.2037037# Simplify a model
# fit_df <- glb_entity_df; glb_mdl <- step(<complex>_mdl)
plot_models_df <- mutate(glb_models_df, feats.label=substr(feats, 1, 30))
if (glb_is_regression) {
print(orderBy(~ -R.sq.OOB -Adj.R.sq.fit, glb_models_df))
stop("glb_sel_mdl not selected")
print(myplot_scatter(plot_models_df, "Adj.R.sq.fit", "R.sq.OOB") +
geom_text(aes(label=feats.label), data=plot_models_df, color="NavyBlue",
size=3.5, angle=45))
}
if (glb_is_classification) {
# Lower AIC is better
print(tmp_models_df <- orderBy(~ -auc.OOB -accuracy.fit +accuracySD.fit, glb_models_df))
print("Selected model:glb_sel_mdl:")
glb_sel_mdl <- glb_models_lst[[as.numeric(rownames(tmp_models_df)[1])]]
print(summary(glb_sel_mdl))
plot_models_df[, "inv.AIC.fit"] <- 1.0 / plot_models_df[, "AIC.fit"]
print(myplot_scatter(plot_models_df, "inv.AIC.fit", "auc.OOB") +
geom_text(aes(label=feats.label), data=plot_models_df, color="NavyBlue",
size=3.5, angle=45))
print(myplot_scatter(plot_models_df, "auc.OOB", "accuracy.fit",
colorcol_name="method") +
geom_errorbar(aes(x=auc.OOB, ymin=accuracy.fit - accuracySD.fit,
ymax=accuracy.fit + accuracySD.fit),
width=(max(plot_models_df$auc.OOB)-min(plot_models_df$auc.OOB))/25) +
geom_text(aes(label=feats.label), data=plot_models_df, color="NavyBlue",
size=3.5, angle=45))
}## method
## 7 rf
## 4 glm
## 5 glm
## 6 rpart
## 2 glm
## 3 glm
## 1 glm
## feats
## 7 Unconst, Circuit.fctr, Issue.fctr, Petitioner.fctr, Respondent.fctr, LowerCourt.fctr
## 4 Term, Unconst, Circuit.fctr, Issue.fctr, Petitioner.fctr, Respondent.fctr, LowerCourt.fctr
## 5 Unconst, Circuit.fctr, Issue.fctr, Petitioner.fctr, Respondent.fctr, LowerCourt.fctr
## 6 Unconst, Circuit.fctr, Issue.fctr, Petitioner.fctr, Respondent.fctr, LowerCourt.fctr
## 2 Unconst
## 3 Unconst, Term
## 1 .rnorm
## n.fit R.sq.fit R.sq.OOB Adj.R.sq.fit SSE.fit SSE.OOB AIC.fit auc.fit
## 7 396 NA NA NA 0.000 NA NA 0.8312757
## 4 396 NA NA NA 2394.341 NA 524.5440 0.7945602
## 5 396 NA NA NA 2419.121 NA 522.7144 0.7948431
## 6 396 NA NA NA 0.000 NA NA 0.6584105
## 2 396 NA NA NA 1604.777 NA 547.9432 0.5277778
## 3 396 NA NA NA 1606.558 NA 549.5810 0.5427083
## 1 396 NA NA NA 1597.666 NA 549.5757 0.5175926
## auc.OOB accuracy.fit accuracySD.fit
## 7 0.7420751 0.6538462 0.060549689
## 4 0.7304846 0.6105128 0.114868742
## 5 0.7267840 0.6113462 0.072459562
## 6 0.7228739 0.6361538 0.078428526
## 2 0.5842759 0.5453846 0.005958436
## 3 0.5704511 0.5403205 0.026811684
## 1 0.4197738 0.5402564 0.015207912
## [1] "Selected model:glb_sel_mdl:"
## Length Class Mode
## call 4 -none- call
## type 1 -none- character
## predicted 396 factor numeric
## err.rate 1500 -none- numeric
## confusion 6 -none- numeric
## votes 792 matrix numeric
## oob.times 396 -none- numeric
## classes 2 -none- character
## importance 46 -none- numeric
## importanceSD 0 -none- NULL
## localImportance 0 -none- NULL
## proximity 0 -none- NULL
## ntree 1 -none- numeric
## mtry 1 -none- numeric
## forest 14 -none- list
## y 396 factor numeric
## test 0 -none- NULL
## inbag 0 -none- NULL
## xNames 46 -none- character
## problemType 1 -none- character
## tuneValue 1 data.frame list
## obsLevels 2 -none- character## Warning: Removed 2 rows containing missing values (geom_point).## Warning: Removed 2 rows containing missing values (geom_text).
replay.petrisim(pn=glb_analytics_pn,
replay.trans=(glb_analytics_avl_objs <- c(glb_analytics_avl_objs,
"model.selected")), flip_coord=TRUE)## time trans "bgn " "fit.data.training.all " "predict.data.new " "end "
## 0.0000 multiple enabled transitions: data.training.all data.new model.selected firing: data.training.all
## 1.0000 1 2 1 0 0
## 1.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction firing: data.new
## 2.0000 2 1 1 1 0
## 2.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction data.new.prediction firing: model.selected
## 3.0000 3 0 2 1 0
glb_script_df <- rbind(glb_script_df,
data.frame(chunk_label="fit.data.training.all",
chunk_step_major=max(glb_script_df$chunk_step_major)+1,
chunk_step_minor=0,
elapsed=(proc.time() - glb_script_tm)["elapsed"]))
print(tail(glb_script_df, 2))## chunk_label chunk_step_major chunk_step_minor elapsed
## elapsed8 run_models 5 0 2.777
## elapsed9 fit.data.training.all 6 0 48.782Step 6: fit.data.training.all
print(mdl_feats_df <- myextract_mdl_feats(lcl_sel_mdl=glb_sel_mdl,
lcl_entity_df=glb_entity_df))## id importance
## LowerCourt.fctr Petitioner.fctr 100.0000000
## Unconst Unconst 19.2694009
## Respondent.fctr Issue.fctr 6.2600400
## Circuit.fctr LowerCourt.fctr 5.3118057
## Issue.fctr Respondent.fctr 0.5524182
## Petitioner.fctr Circuit.fctr 0.0000000ret_lst <- myrun_mdl_fn(indep_vars_vctr=mdl_feats_df$id,
lcl_predct_var=glb_predct_var,
lcl_predct_var_name=glb_predct_var_name,
fit_df=glb_entity_df,
method=glb_sel_mdl$method,
lcl_tune_models_df=glb_tune_models_df)
##
## Call:
## randomForest(x = x, y = y, mtry = param$mtry)
## Type of random forest: classification
## Number of trees: 500
## No. of variables tried at each split: 3
##
## OOB estimate of error rate: 36.87%
## Confusion matrix:
## A R class.error
## A 90 90 0.5000000
## R 56 160 0.2592593glb_sel_mdl <- ret_lst[["model"]]; #glb_sel_mdl_fn <- ret_lst[["model_fn"]]
# print(glb_models_df[nrow(glb_models_df), ])
glb_script_df <- rbind(glb_script_df,
data.frame(chunk_label="fit.data.training.all",
chunk_step_major=glb_script_df[nrow(glb_script_df), "chunk_step_major"],
chunk_step_minor=glb_script_df[nrow(glb_script_df), "chunk_step_minor"]+1,
elapsed=(proc.time() - glb_script_tm)["elapsed"]))
print(tail(glb_script_df, 2))## chunk_label chunk_step_major chunk_step_minor elapsed
## elapsed9 fit.data.training.all 6 0 48.782
## elapsed10 fit.data.training.all 6 1 71.905if (glb_is_regression) {
glb_entity_df[, glb_predct_var_name] <- predict(glb_sel_mdl, newdata=glb_entity_df)
print(myplot_scatter(glb_entity_df, glb_predct_var, glb_predct_var_name,
smooth=TRUE))
glb_entity_df[, paste0(glb_predct_var_name, ".err")] <-
abs(glb_entity_df[, glb_predct_var_name] - glb_entity_df[, glb_predct_var])
print(head(orderBy(reformulate(c("-", paste0(glb_predct_var_name, ".err"))),
glb_entity_df)))
}
if (glb_is_classification) {
if (any(class(glb_sel_mdl) %in% c("train"))) {
glb_entity_df[, paste0(glb_predct_var_name, ".proba")] <-
predict(glb_sel_mdl, newdata=glb_entity_df, type="prob")[, 2]
} else if (any(class(glb_sel_mdl) %in% c("rpart", "randomForest"))) {
glb_entity_df[, paste0(glb_predct_var_name, ".proba")] <-
predict(glb_sel_mdl, newdata=glb_entity_df, type="prob")[, 2]
} else if (class(glb_sel_mdl) == "glm") {
stop("not implemented yet")
glb_entity_df[, paste0(glb_predct_var_name, ".proba")] <-
predict(glb_sel_mdl, newdata=glb_entity_df, type="response")
} else stop("not implemented yet")
require(ROCR)
ROCRpred <- prediction(glb_entity_df[, paste0(glb_predct_var_name, ".proba")],
glb_entity_df[, glb_predct_var])
ROCRperf <- performance(ROCRpred, "tpr", "fpr")
plot(ROCRperf, colorize=TRUE, print.cutoffs.at=seq(0, 1, 0.1), text.adj=c(-0.2,1.7))
thresholds_df <- data.frame(threshold=seq(0.0, 1.0, 0.1))
thresholds_df$f.score <- sapply(1:nrow(thresholds_df), function(row_ix)
mycompute_classifier_f.score(mdl=glb_sel_mdl, obs_df=glb_entity_df,
proba_threshold=thresholds_df[row_ix, "threshold"],
lcl_predct_var=glb_predct_var,
lcl_predct_var_name=glb_predct_var_name))
print(thresholds_df)
print(myplot_line(thresholds_df, "threshold", "f.score"))
proba_threshold <- thresholds_df[which.max(thresholds_df$f.score),
"threshold"]
# This should change to maximize f.score.OOB ???
print(sprintf("Classifier Probability Threshold: %0.4f to maximize f.score.fit",
proba_threshold))
if (is.null(glb_clf_proba_threshold))
glb_clf_proba_threshold <- proba_threshold else {
print(sprintf("Classifier Probability Threshold: %0.4f per user specs",
glb_clf_proba_threshold))
}
if ((class(glb_entity_df[, glb_predct_var]) != "factor") |
(length(levels(glb_entity_df[, glb_predct_var])) != 2))
stop("expecting a factor with two levels:", glb_predct_var)
glb_entity_df[, glb_predct_var_name] <-
factor(levels(glb_entity_df[, glb_predct_var])[
(glb_entity_df[, paste0(glb_predct_var_name, ".proba")] >=
glb_clf_proba_threshold) * 1 + 1])
print(mycreate_xtab(glb_entity_df, c(glb_predct_var, glb_predct_var_name)))
print(sprintf("f.score=%0.4f",
mycompute_classifier_f.score(glb_sel_mdl, glb_entity_df,
glb_clf_proba_threshold,
glb_predct_var, glb_predct_var_name)))
} 
## threshold f.score
## 1 0.0 0.7058824
## 2 0.1 0.7200000
## 3 0.2 0.7544484
## 4 0.3 0.7842402
## 5 0.4 0.7967480
## 6 0.5 0.8219780
## 7 0.6 0.8000000
## 8 0.7 0.7022472
## 9 0.8 0.5833333
## 10 0.9 0.2650602
## 11 1.0 0.0000000
## [1] "Classifier Probability Threshold: 0.5000 to maximize f.score.fit"
## [1] "Classifier Probability Threshold: 0.5000 per user specs"
## Decision.fctr Decision.fctr.predict.A Decision.fctr.predict.R
## 1 A 128 52
## 2 R 29 187
## [1] "f.score=0.8220"print(glb_feats_df <- mymerge_feats_importance(glb_feats_df, glb_sel_mdl, glb_entity_df))## id cor.y cor.y.abs cor.low importance
## 4 Petitioner.fctr NA NA NA 100.0000000
## 7 Unconst 0.06627144 0.06627144 1 25.4763117
## 2 Issue.fctr NA NA NA 7.2530286
## 3 LowerCourt.fctr NA NA NA 5.4965117
## 5 Respondent.fctr NA NA NA 0.6156672
## 1 Circuit.fctr NA NA NA 0.0000000
## 6 Term 0.03032304 0.03032304 1 NA# Most of this code is used again in predict.data.new chunk
glb_analytics_diag_plots <- function(obs_df) {
for (var in subset(glb_feats_df, !is.na(importance))$id) {
plot_df <- melt(obs_df, id.vars=var,
measure.vars=c(glb_predct_var, glb_predct_var_name))
# if (var == "<feat_name>") print(myplot_scatter(plot_df, var, "value",
# facet_colcol_name="variable") +
# geom_vline(xintercept=<divider_val>, linetype="dotted")) else
print(myplot_scatter(plot_df, var, "value",
facet_colcol_name="variable", jitter=TRUE))
}
if (glb_is_regression) {
plot_vars_df <- subset(glb_feats_df, Pr.z < 0.1)
print(myplot_prediction_regression(obs_df,
ifelse(nrow(plot_vars_df) > 1, plot_vars_df$id[2], ".rownames"),
plot_vars_df$id[1],
glb_predct_var, glb_predct_var_name)
# + facet_wrap(reformulate(plot_vars_df$id[2])) # if [1,2] is a factor
# + geom_point(aes_string(color="<col_name>.fctr")) # to color the plot
)
}
if (glb_is_classification) {
if (nrow(plot_vars_df <- subset(glb_feats_df, !is.na(importance))) == 0)
warning("No features in selected model are statistically important")
else print(myplot_prediction_classification(df=obs_df,
feat_x=ifelse(nrow(plot_vars_df) > 1, plot_vars_df$id[2],
".rownames"),
feat_y=plot_vars_df$id[1],
lcl_predct_var=glb_predct_var,
lcl_predct_var_name=glb_predct_var_name,
lcl_id_vars=glb_id_vars)
# + geom_hline(yintercept=<divider_val>, linetype = "dotted")
)
}
}
glb_analytics_diag_plots(obs_df=glb_entity_df)
## Docket Term Circuit Issue Petitioner Respondent
## 2 93-1577 1994 9th EconomicActivity BUSINESS BUSINESS
## 12 99-1571 2000 6th EconomicActivity BUSINESS BUSINESS
## 20 95-1872 1996 8th CivilRights AMERICAN.INDIAN BUSINESS
## 28 00-152 2000 9th JudicialPower GOVERNMENT.OFFICIAL BUSINESS
## 35 94-1592 1995 7th Unions OTHER BUSINESS
## 37 95-26 1995 FED EconomicActivity OTHER BUSINESS
## LowerCourt Unconst Reverse Reverse.fctr Decision.fctr Circuit.fctr
## 2 liberal 0 1 1 R 9th
## 12 conser 1 1 1 R 6th
## 20 conser 0 0 0 A 8th
## 28 liberal 0 1 1 R 9th
## 35 conser 0 0 0 A 7th
## 37 conser 1 0 0 A FED
## Issue.fctr Petitioner.fctr Respondent.fctr LowerCourt.fctr
## 2 EconomicActivity BUSINESS BUSINESS liberal
## 12 EconomicActivity BUSINESS BUSINESS conser
## 20 CivilRights AMERICAN.INDIAN BUSINESS conser
## 28 JudicialPower GOVERNMENT.OFFICIAL BUSINESS liberal
## 35 Unions OTHER BUSINESS conser
## 37 EconomicActivity OTHER BUSINESS conser
## .rnorm Decision.fctr.predict.proba Decision.fctr.predict
## 2 0.3156464 0.484 A
## 12 0.2705301 0.834 R
## 20 1.5200465 0.568 R
## 28 -0.3929619 0.214 A
## 35 -0.3006087 0.652 R
## 37 -0.4273334 0.510 R
## Decision.fctr.fctr Decision.fctr.predict.accurate .label
## 2 R FALSE 93-1577
## 12 R TRUE 99-1571
## 20 A FALSE 95-1872
## 28 R FALSE 00-152
## 35 A FALSE 94-1592
## 37 A FALSE 95-26
## Docket Term Circuit Issue Petitioner
## 12 99-1571 2000 6th EconomicActivity BUSINESS
## 20 95-1872 1996 8th CivilRights AMERICAN.INDIAN
## 61 00-276 2000 6th EconomicActivity OTHER
## 200 94-771 1994 10th CivilRights OTHER
## 270 98-84 1998 3rd CivilRights BUSINESS
## 529 97-7164 1998 2nd CriminalProcedure CRIMINAL.DEFENDENT
## Respondent LowerCourt Unconst Reverse Reverse.fctr Decision.fctr
## 12 BUSINESS conser 1 1 1 R
## 20 BUSINESS conser 0 0 0 A
## 61 BUSINESS liberal 0 0 0 A
## 200 AMERICAN.INDIAN liberal 0 0 0 A
## 270 OTHER liberal 0 1 1 R
## 529 US conser 0 0 0 A
## Circuit.fctr Issue.fctr Petitioner.fctr Respondent.fctr
## 12 6th EconomicActivity BUSINESS BUSINESS
## 20 8th CivilRights AMERICAN.INDIAN BUSINESS
## 61 6th EconomicActivity OTHER BUSINESS
## 200 10th CivilRights OTHER AMERICAN.INDIAN
## 270 3rd CivilRights BUSINESS OTHER
## 529 2nd CriminalProcedure CRIMINAL.DEFENDENT US
## LowerCourt.fctr .rnorm Decision.fctr.predict.proba
## 12 conser 0.2705301 0.834
## 20 conser 1.5200465 0.568
## 61 liberal 0.9395453 0.656
## 200 liberal -0.5603275 0.670
## 270 liberal -0.2974204 0.426
## 529 conser -1.1304355 0.696
## Decision.fctr.predict Decision.fctr.fctr
## 12 R R
## 20 R A
## 61 R A
## 200 R A
## 270 A R
## 529 R A
## Decision.fctr.predict.accurate .label
## 12 TRUE 99-1571
## 20 FALSE 95-1872
## 61 FALSE 00-276
## 200 FALSE 94-771
## 270 FALSE 98-84
## 529 FALSE 97-7164
## Docket Term Circuit Issue Petitioner Respondent
## 529 97-7164 1998 2nd CriminalProcedure CRIMINAL.DEFENDENT US
## 532 97-9217 1998 3rd CriminalProcedure CRIMINAL.DEFENDENT US
## 542 99-9073 2000 7th CriminalProcedure CRIMINAL.DEFENDENT US
## 557 96-262 1996 FED JudicialPower OTHER US
## 562 99-5153 1999 6th CriminalProcedure OTHER US
## 564 99-8508 2000 9th CriminalProcedure OTHER US
## LowerCourt Unconst Reverse Reverse.fctr Decision.fctr Circuit.fctr
## 529 conser 0 0 0 A 2nd
## 532 conser 0 0 0 A 3rd
## 542 conser 0 0 0 A 7th
## 557 conser 0 0 0 A FED
## 562 conser 0 0 0 A 6th
## 564 conser 1 0 0 A 9th
## Issue.fctr Petitioner.fctr Respondent.fctr LowerCourt.fctr
## 529 CriminalProcedure CRIMINAL.DEFENDENT US conser
## 532 CriminalProcedure CRIMINAL.DEFENDENT US conser
## 542 CriminalProcedure CRIMINAL.DEFENDENT US conser
## 557 JudicialPower OTHER US conser
## 562 CriminalProcedure OTHER US conser
## 564 CriminalProcedure OTHER US conser
## .rnorm Decision.fctr.predict.proba Decision.fctr.predict
## 529 -1.1304355 0.696 R
## 532 -0.9045173 0.572 R
## 542 0.8678893 0.568 R
## 557 1.6463979 0.566 R
## 562 0.4310960 0.522 R
## 564 -0.1638353 0.560 R
## Decision.fctr.fctr Decision.fctr.predict.accurate .label
## 529 A FALSE 97-7164
## 532 A FALSE 97-9217
## 542 A FALSE 99-9073
## 557 A FALSE 96-262
## 562 A FALSE 99-5153
## 564 A FALSE 99-8508
replay.petrisim(pn=glb_analytics_pn,
replay.trans=(glb_analytics_avl_objs <- c(glb_analytics_avl_objs,
"data.training.all.prediction","model.final")), flip_coord=TRUE)## time trans "bgn " "fit.data.training.all " "predict.data.new " "end "
## 0.0000 multiple enabled transitions: data.training.all data.new model.selected firing: data.training.all
## 1.0000 1 2 1 0 0
## 1.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction firing: data.new
## 2.0000 2 1 1 1 0
## 2.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction data.new.prediction firing: model.selected
## 3.0000 3 0 2 1 0
## 3.0000 multiple enabled transitions: model.final data.training.all.prediction data.new.prediction firing: data.training.all.prediction
## 4.0000 5 0 1 1 1
## 4.0000 multiple enabled transitions: model.final data.training.all.prediction data.new.prediction firing: model.final
## 5.0000 4 0 0 2 1
glb_script_df <- rbind(glb_script_df,
data.frame(chunk_label="predict.data.new",
chunk_step_major=max(glb_script_df$chunk_step_major)+1,
chunk_step_minor=0,
elapsed=(proc.time() - glb_script_tm)["elapsed"]))
print(tail(glb_script_df, 2))## chunk_label chunk_step_major chunk_step_minor elapsed
## elapsed10 fit.data.training.all 6 1 71.905
## elapsed11 predict.data.new 7 0 76.635Step 7: predict data.new
if (glb_is_regression)
glb_newent_df[, glb_predct_var_name] <- predict(glb_sel_mdl,
newdata=glb_newent_df, type="response")
if (glb_is_classification) {
# Compute selected model predictions
if (any(class(glb_sel_mdl) %in% c("train"))) {
glb_newent_df[, paste0(glb_predct_var_name, ".proba")] <-
predict(glb_sel_mdl, newdata=glb_newent_df, type="prob")[, 2]
} else if (any(class(glb_sel_mdl) %in% c("rpart", "randomForest"))) {
glb_newent_df[, paste0(glb_predct_var_name, ".proba")] <-
predict(glb_sel_mdl, newdata=glb_newent_df, type="prob")[, 2]
} else if (class(glb_sel_mdl) == "glm") {
stop("not implemented yet")
glb_newent_df[, paste0(glb_predct_var_name, ".proba")] <-
predict(glb_sel_mdl, newdata=glb_newent_df, type="response")
} else stop("not implemented yet")
if ((class(glb_newent_df[, glb_predct_var]) != "factor") |
(length(levels(glb_newent_df[, glb_predct_var])) != 2))
stop("expecting a factor with two levels:", glb_predct_var)
glb_newent_df[, glb_predct_var_name] <-
factor(levels(glb_newent_df[, glb_predct_var])[
(glb_newent_df[, paste0(glb_predct_var_name, ".proba")] >=
glb_clf_proba_threshold) * 1 + 1])
# Compute dummy model predictions
glb_newent_df[, paste0(glb_predct_var, ".preddmy.proba")] <-
predict(glb_dmy_mdl, newdata=glb_newent_df, type="prob")[, 2]
if ((class(glb_newent_df[, glb_predct_var]) != "factor") |
(length(levels(glb_newent_df[, glb_predct_var])) != 2))
stop("expecting a factor with two levels:", glb_predct_var)
glb_newent_df[, paste0(glb_predct_var, ".preddmy")] <-
factor(levels(glb_newent_df[, glb_predct_var])[
(glb_newent_df[, paste0(glb_predct_var, ".preddmy.proba")] >=
glb_clf_proba_threshold) * 1 + 1])
}
myprint_df(glb_newent_df[, c(glb_id_vars, glb_predct_var, glb_predct_var_name)])## Docket Decision.fctr Decision.fctr.predict
## 1 93-1408 R R
## 3 93-1612 R R
## 4 94-623 R R
## 6 95-129 A R
## 8 96-1768 R R
## 21 97-1704 A R
## Docket Decision.fctr Decision.fctr.predict
## 47 96-1470 R R
## 268 97-303 A A
## 350 95-938 R R
## 432 99-1864 R R
## 478 00-5961 R R
## 480 98-942 R R
## Docket Decision.fctr Decision.fctr.predict
## 545 00-507 A R
## 546 95-1478 A A
## 551 00-157 A A
## 552 00-1567 A A
## 556 95-173 R R
## 558 96-6839 R Rif (glb_is_regression) {
print(sprintf("Total SSE: %0.4f",
sum((glb_newent_df[, glb_predct_var_name] -
glb_newent_df[, glb_predct_var]) ^ 2)))
print(sprintf("RMSE: %0.4f",
(sum((glb_newent_df[, glb_predct_var_name] -
glb_newent_df[, glb_predct_var]) ^ 2) / nrow(glb_newent_df)) ^ 0.5))
print(myplot_scatter(glb_newent_df, glb_predct_var, glb_predct_var_name,
smooth=TRUE))
glb_newent_df[, paste0(glb_predct_var_name, ".err")] <-
abs(glb_newent_df[, glb_predct_var_name] - glb_newent_df[, glb_predct_var])
print(head(orderBy(reformulate(c("-", paste0(glb_predct_var_name, ".err"))),
glb_newent_df)))
# glb_newent_df[, "<Output Pred variable>"] <- func(glb_newent_df[, glb_pred_var_name])
}
if (glb_is_classification) {
ROCRpred <- prediction(glb_newent_df[, paste0(glb_predct_var_name, ".proba")],
glb_newent_df[, glb_predct_var])
print(sprintf("auc=%0.4f", auc <- as.numeric(performance(ROCRpred, "auc")@y.values)))
print(sprintf("probability threshold=%0.4f", glb_clf_proba_threshold))
print(newent_conf_df <- mycreate_xtab(glb_newent_df,
c(glb_predct_var, glb_predct_var_name)))
print(sprintf("f.score.sel=%0.4f",
mycompute_classifier_f.score(mdl=glb_sel_mdl, obs_df=glb_newent_df,
proba_threshold=glb_clf_proba_threshold,
lcl_predct_var=glb_predct_var,
lcl_predct_var_name=glb_predct_var_name)))
print(sprintf("sensitivity=%0.4f", newent_conf_df[2, 3] /
(newent_conf_df[2, 3] + newent_conf_df[2, 2])))
print(sprintf("specificity=%0.4f", newent_conf_df[1, 2] /
(newent_conf_df[1, 2] + newent_conf_df[1, 3])))
print(sprintf("accuracy=%0.4f", (newent_conf_df[1, 2] + newent_conf_df[2, 3]) /
(newent_conf_df[1, 2] + newent_conf_df[2, 3] +
newent_conf_df[1, 3] + newent_conf_df[2, 2])))
print(mycreate_xtab(glb_newent_df, c(glb_predct_var, paste0(glb_predct_var, ".preddmy"))))
print(sprintf("f.score.dmy=%0.4f",
mycompute_classifier_f.score(mdl=glb_dmy_mdl, obs_df=glb_newent_df,
proba_threshold=glb_clf_proba_threshold,
lcl_predct_var=glb_predct_var,
lcl_predct_var_name=paste0(glb_predct_var, ".preddmy"))))
} ## [1] "auc=0.7366"
## [1] "probability threshold=0.5000"
## Decision.fctr Decision.fctr.predict.A Decision.fctr.predict.R
## 1 A 44 33
## 2 R 21 72
## [1] "f.score.sel=0.7273"
## [1] "sensitivity=0.7742"
## [1] "specificity=0.5714"
## [1] "accuracy=0.6824"
## Decision.fctr Decision.fctr.preddmy.R
## 1 A 77
## 2 R 93
## [1] "f.score.dmy=0.7072"glb_analytics_diag_plots(glb_newent_df)
## Docket Term Circuit Issue Petitioner Respondent
## 1 93-1408 1994 2nd EconomicActivity BUSINESS BUSINESS
## 6 95-129 1995 9th EconomicActivity BUSINESS BUSINESS
## 21 97-1704 1998 5th JudicialPower INJURED.PERSON BUSINESS
## 40 95-1065 1996 2nd FirstAmendment OTHER BUSINESS
## 53 97-1130 1998 FED EconomicActivity OTHER BUSINESS
## 66 99-1871 2000 9th Privacy OTHER BUSINESS
## LowerCourt Unconst Reverse Reverse.fctr Decision.fctr Circuit.fctr
## 1 liberal 0 1 1 R 2nd
## 6 conser 1 0 0 A 9th
## 21 liberal 1 0 0 A 5th
## 40 liberal 1 0 0 A 2nd
## 53 liberal 0 0 0 A FED
## 66 conser 0 0 0 A 9th
## Issue.fctr Petitioner.fctr Respondent.fctr LowerCourt.fctr
## 1 EconomicActivity BUSINESS BUSINESS liberal
## 6 EconomicActivity BUSINESS BUSINESS conser
## 21 JudicialPower INJURED.PERSON BUSINESS liberal
## 40 FirstAmendment OTHER BUSINESS liberal
## 53 EconomicActivity OTHER BUSINESS liberal
## 66 Privacy OTHER BUSINESS conser
## .rnorm Decision.fctr.predict.proba Decision.fctr.predict
## 1 -0.6413986 0.690 R
## 6 -0.8008941 0.756 R
## 21 -0.4487406 0.528 R
## 40 -0.3238044 0.512 R
## 53 -0.4352795 0.572 R
## 66 0.9040435 0.918 R
## Decision.fctr.preddmy.proba Decision.fctr.preddmy Decision.fctr.fctr
## 1 0.5514549 R R
## 6 0.5527867 R A
## 21 0.5498453 R A
## 40 0.5488009 R A
## 53 0.5497328 R A
## 66 0.5385155 R A
## Decision.fctr.predict.accurate .label
## 1 TRUE 93-1408
## 6 FALSE 95-129
## 21 FALSE 97-1704
## 40 FALSE 95-1065
## 53 FALSE 97-1130
## 66 FALSE 99-1871
## Docket Term Circuit Issue Petitioner Respondent
## 66 99-1871 2000 9th Privacy OTHER BUSINESS
## 148 94-325 1994 2nd EconomicActivity BUSINESS EMPLOYEE
## 175 96-651 1996 3rd Unions STATE EMPLOYEE
## 394 99-603 2000 2nd FirstAmendment OTHER OTHER
## 441 94-558 1994 5th CivilRights US POLITICIAN
## 482 99-1964 2000 3rd JudicialPower CRIMINAL.DEFENDENT STATE
## LowerCourt Unconst Reverse Reverse.fctr Decision.fctr Circuit.fctr
## 66 conser 0 0 0 A 9th
## 148 liberal 0 0 0 A 2nd
## 175 liberal 0 1 1 R 3rd
## 394 liberal 0 0 0 A 2nd
## 441 liberal 0 1 1 R 5th
## 482 conser 0 0 0 A 3rd
## Issue.fctr Petitioner.fctr Respondent.fctr LowerCourt.fctr
## 66 Privacy OTHER BUSINESS conser
## 148 EconomicActivity BUSINESS EMPLOYEE liberal
## 175 Unions STATE EMPLOYEE liberal
## 394 FirstAmendment OTHER OTHER liberal
## 441 CivilRights US POLITICIAN liberal
## 482 JudicialPower CRIMINAL.DEFENDENT STATE conser
## .rnorm Decision.fctr.predict.proba Decision.fctr.predict
## 66 0.9040435 0.918 R
## 148 -0.2926154 0.584 R
## 175 0.4397330 0.202 A
## 394 0.5146458 0.594 R
## 441 -0.4416542 0.458 A
## 482 1.3568613 0.672 R
## Decision.fctr.preddmy.proba Decision.fctr.preddmy Decision.fctr.fctr
## 66 0.5385155 R A
## 148 0.5485401 R A
## 175 0.5424092 R R
## 394 0.5417813 R A
## 441 0.5497861 R R
## 482 0.5347135 R A
## Decision.fctr.predict.accurate .label
## 66 FALSE 99-1871
## 148 FALSE 94-325
## 175 FALSE 96-651
## 394 FALSE 99-603
## 441 FALSE 94-558
## 482 FALSE 99-1964
## Docket Term Circuit Issue Petitioner Respondent
## 468 98-5864 1998 4th CriminalProcedure CRIMINAL.DEFENDENT STATE
## 482 99-1964 2000 3rd JudicialPower CRIMINAL.DEFENDENT STATE
## 494 99-2047 2000 1st CivilRights OTHER STATE
## 504 94-1664 1995 9th CriminalProcedure CRIMINAL.DEFENDENT US
## 527 97-6270 1997 1st CriminalProcedure CRIMINAL.DEFENDENT US
## 545 00-507 2001 10th CivilRights AMERICAN.INDIAN US
## LowerCourt Unconst Reverse Reverse.fctr Decision.fctr Circuit.fctr
## 468 conser 1 0 0 A 4th
## 482 conser 0 0 0 A 3rd
## 494 liberal 1 0 0 A 1st
## 504 conser 0 0 0 A 9th
## 527 conser 0 0 0 A 1st
## 545 conser 0 0 0 A 10th
## Issue.fctr Petitioner.fctr Respondent.fctr LowerCourt.fctr
## 468 CriminalProcedure CRIMINAL.DEFENDENT STATE conser
## 482 JudicialPower CRIMINAL.DEFENDENT STATE conser
## 494 CivilRights OTHER STATE liberal
## 504 CriminalProcedure CRIMINAL.DEFENDENT US conser
## 527 CriminalProcedure CRIMINAL.DEFENDENT US conser
## 545 CivilRights AMERICAN.INDIAN US conser
## .rnorm Decision.fctr.predict.proba Decision.fctr.predict
## 468 -0.536313833 0.990 R
## 482 1.356861338 0.672 R
## 494 -0.221110732 0.532 R
## 504 1.236920941 0.732 R
## 527 -0.003943005 0.678 R
## 545 0.222748927 0.510 R
## Decision.fctr.preddmy.proba Decision.fctr.preddmy Decision.fctr.fctr
## 468 0.5505771 R A
## 482 0.5347135 R A
## 494 0.5479421 R A
## 504 0.5357210 R A
## 527 0.5461251 R A
## 545 0.5442272 R A
## Decision.fctr.predict.accurate .label
## 468 FALSE 98-5864
## 482 FALSE 99-1964
## 494 FALSE 99-2047
## 504 FALSE 94-1664
## 527 FALSE 97-6270
## 545 FALSE 00-507
tmp_replay_lst <- replay.petrisim(pn=glb_analytics_pn,
replay.trans=(glb_analytics_avl_objs <- c(glb_analytics_avl_objs,
"data.new.prediction")), flip_coord=TRUE)## time trans "bgn " "fit.data.training.all " "predict.data.new " "end "
## 0.0000 multiple enabled transitions: data.training.all data.new model.selected firing: data.training.all
## 1.0000 1 2 1 0 0
## 1.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction firing: data.new
## 2.0000 2 1 1 1 0
## 2.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction data.new.prediction firing: model.selected
## 3.0000 3 0 2 1 0
## 3.0000 multiple enabled transitions: model.final data.training.all.prediction data.new.prediction firing: data.training.all.prediction
## 4.0000 5 0 1 1 1
## 4.0000 multiple enabled transitions: model.final data.training.all.prediction data.new.prediction firing: model.final
## 5.0000 4 0 0 2 1
## 6.0000 6 0 0 1 2
#print(ggplot.petrinet(tmp_replay_lst[["pn"]]) + coord_flip())Null Hypothesis (\(\sf{H_{0}}\)): mpg is not impacted by am_fctr.
The variance by am_fctr appears to be independent.
# print(t.test(subset(cars_df, am_fctr == "automatic")$mpg,
# subset(cars_df, am_fctr == "manual")$mpg,
# var.equal=FALSE)$conf)We reject the null hypothesis i.e. we have evidence to conclude that am_fctr impacts mpg (95% confidence). Manual transmission is better for miles per gallon versus automatic transmission.
## chunk_label chunk_step_major chunk_step_minor elapsed
## 10 fit.data.training.all 6 0 48.782
## 11 fit.data.training.all 6 1 71.905
## 12 predict.data.new 7 0 76.635
## 4 manage_missing_data 2 2 1.428
## 7 select_features 4 0 2.526
## 2 cleanse_data 2 0 0.564
## 5 encode_retype_data 2 3 1.668
## 8 remove_correlated_features 4 1 2.710
## 9 run_models 5 0 2.777
## 6 extract_features 3 0 1.715
## 3 inspect_explore_data 2 1 0.595
## 1 import_data 1 0 0.002
## elapsed_diff
## 10 46.005
## 11 23.123
## 12 4.730
## 4 0.833
## 7 0.811
## 2 0.562
## 5 0.240
## 8 0.184
## 9 0.067
## 6 0.047
## 3 0.031
## 1 0.000## [1] "Total Elapsed Time: 76.635 secs"
## R version 3.1.3 (2015-03-09)
## Platform: x86_64-apple-darwin13.4.0 (64-bit)
## Running under: OS X 10.10.2 (Yosemite)
##
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
##
## attached base packages:
## [1] grid stats graphics grDevices utils datasets methods
## [8] base
##
## other attached packages:
## [1] randomForest_4.6-10 rpart.plot_1.5.2 rpart_4.1-9
## [4] caret_6.0-41 lattice_0.20-30 ROCR_1.0-6
## [7] gplots_2.16.0 reshape2_1.4.1 plyr_1.8.1
## [10] caTools_1.17.1 doBy_4.5-13 survival_2.38-1
## [13] ggplot2_1.0.1
##
## loaded via a namespace (and not attached):
## [1] bitops_1.0-6 BradleyTerry2_1.0-6 brglm_0.5-9
## [4] car_2.0-25 class_7.3-12 codetools_0.2-10
## [7] colorspace_1.2-6 compiler_3.1.3 digest_0.6.8
## [10] e1071_1.6-4 evaluate_0.5.5 foreach_1.4.2
## [13] formatR_1.0 gdata_2.13.3 gtable_0.1.2
## [16] gtools_3.4.1 htmltools_0.2.6 iterators_1.0.7
## [19] KernSmooth_2.23-14 knitr_1.9 labeling_0.3
## [22] lme4_1.1-7 MASS_7.3-39 Matrix_1.1-5
## [25] mgcv_1.8-4 minqa_1.2.4 munsell_0.4.2
## [28] nlme_3.1-120 nloptr_1.0.4 nnet_7.3-9
## [31] parallel_3.1.3 pbkrtest_0.4-2 proto_0.3-10
## [34] quantreg_5.11 Rcpp_0.11.5 rmarkdown_0.5.1
## [37] scales_0.2.4 SparseM_1.6 splines_3.1.3
## [40] stringr_0.6.2 tools_3.1.3 yaml_2.1.13